Post Event Compensatory Trends

The deeper analysis of our data yields surprises, and sometimes even requires re-examination of notions built up over decades of research at the edges of what we know about consciousness. Peter Bancel has been looking at the timecourse of deviations associated with the formal events. One aspect of the picture is a somewhat surprising, albeit subtle suggestion that "what goes up must come down." The most recent (mid-2007) analyses addressing this issue show clear indications of a kind of "recovery" phase in addition to the original effect on the data associated with major events. At first glance this looks like there must be some strange tendency for nature to remember the deviations from randomness and compensate for them. But such an accounting is not in the nature of mother nature, or more specifically, it is not in the nature of randomness. The expectation for a random sequence unaffected by some influence is that the next element is unpredictable. That implies that wherever the sequence is now is always the starting point for the future sequence -- there is no influence of past history.

But the remarkable findings of this analysis, shown in the graphs below, do require some explanation. If it isn't the world pushing back against an anomaly, what might cause this symmetrical reversal of trend? I think the most likely explanation is differences in the state or quality of "global consciousness" over the time course of a major event. It isn't hard to imagine that at the beginning of a major event its newness and our surprise might dominate our presence, and as we understand more or progress in our experience, the shared perception would also evolve. For example, a terrorist attack will arouse shock and astonishment, and most likely fear at first, perhaps the beginnings of anger. And then we may shift to feelings of deep concern and compassion for those hurt and eventually for their loved ones. Later, the intensity and depth of feeling must transform into profound sadness and a worldly fatigue. As we learn more about the psychological and social context of the effects, and about other possible influences, the picture may become more clear.

The first figure shows the data during the event in blue, and data from the subsequent similar period in red. The measure is a combination of Netvar and Covar which is called the Dispersion Statistic, and the plot shows the cumulative deviation of the statistic from expectation. The data from all events amenable to this analysis (N=187) are used.

Post Event
Compensatory Trends

The second figure presents the same statistics in a simple time sequence. In this case four segments are shown in color to allow easy identification. The first red segment is a pre-event period of the same length as the defined event. The first blue segment is the event itself. Then we see the post event "recovery" period and finally another segment of the same length. Both the first and last segments are unexceptional random walks, but the event and the post-event trends are strongly deviant, and the slopes are of opposite sign. The trend during the event is a little longer, but both are significant.

Post Event
Compensatory Trends


GCP Home